The gunshot sound produced by male North (1)

zyxwvuts
zyxw

MARINE MAMMAL SCIENCE, 21(3):458475 (July 2005)
0 2005 by the Society for Marine Mammalogy

THE GUNSHOT SOUND PRODUCED BY
MALE NORTH ATLANTIC RIGHT
WHALES (EUBALAENA GLACIALIS) AND
ITS POTENTIAL FUNCTION IN
REPRODUCTIVE ADVERTISEMENT
SUSANE. PARKS'
Woods Hole Oceanographic Institution,
Redfield 132, MS 34,
Woods Hole, Massachusetts 02543, U.S.A.
E-mail: sparks@whoi.edu

PHILIPK. HAMILTON
SCOTTD. KRAUS
Edgerton Research Laboratory,
New England Aquarium,

Central Wharf,
Boston, Massachusetts 0 2 1 10, U.S.A.

PETERL. TYACK

zyxwv

Woods Hole Oceanographic Institution,
Redfield 132, MS 34,
Woods Hole, Massachusetts 02543, U.S.A

zyx
zyxwvut
ABSTRACT

North Atlantic right whales (Eubdlaena gZ&alis) make a short, distinctive
broadband sound that is produced internally called a Gunshot sound. This sound has
been recorded in the Bay of Fundy, Canada from both single whales (n =9) and social
surface active groups (n=49). Those single whales producing Gunshot sounds whose
sex could be determined ( n = 9) were all mature males. Gunshot sounds were

produced as part of a stereotyped behavioral sequence by these individuals, including
frequent head-lifts and flipper slapping at the surface. In surface active groups,
Gunshot sounds were commonly recorded when males were present in the group.
The rate of production of Gunshot sounds was weakly correlated with the total
number of males present in the group. Given the behavioral context of Gunshot
sound production, and production of the sound only by male whales, Gunshots may
function in a reproductive context as an advertisement signal to attract females, an
agonistic signal directed toward other males, or a combination of the two functions.
Key words: right whale, Eubalaena glacialis, acoustics, communication, male
advertisement, Gunshot sound, surface active group, Bay of Fundy.

Current mailing address: Bioacoustics Research Program, Cornell Laboratory of Ornithology, 159
Sapsucker Woods Road, Ithaca, New York 14850, U.S.A.

458

zyxwvutsrqp
zyxwvutsr
zyxwvutsrq

PARKS ET AL.: RIGHT WHALE COMMUNICATION

459

Male acoustic advertisement is common in the mating systems of terrestrial
animals where the females are mobile and can seek out mates (Greenwood 1980). In
these species, males often make signals that differ from those of females. In baleen
whales, both males and females are highly mobile and the mating system is not well
understood for any species. The production of distinctive sounds by males has been
documented in several large whale species (Payne and McVay 1971, Tyack 1981,
Weilgart and Whitehead 1988, McDonald et al. 2001, Croll et al. 2002), suggesting
the possibility that males use sound in reproductive advertisement or intrasexual
competition. Little behavioral information exists about sound production related to
reproduction in North Atlantic right whales (Kraus and Hatch 2001, Parks 2003a)
and no information has been published on sounds attributed to males in potentially
reproductive contexts. This paper describes behavioral observations and acoustic
recordings of a short, intense broadband acoustic signal produced by male North
Atlantic right whales (Eubalaenaglacialis)in the Bay of Fundy, Canada. This sound is
referred to as a Gunshot sound because of its acoustic similarity to the sound of a rifle
being fired.

The Gunshot sound has been mentioned in previous publications on balaenid
whale species including Southern right whales (Eubalaena anstralzs) (Clark 1983),
North Atlantic right whales (Matthews et al. 2001, Laurinolli et al. 2003, Parks
200%) and the Bowhead whale (Balaena mysticetns) (Wursig and Clark 1993). In
Southern right whales, Gunshot sounds were recorded both from single whales and
from groups (Clark 1983). The sound has been referred to as a “pulse” (Payne and
Payne 1971), “underwater slap” (Clark 1983), and Gunshot (Clark 1983, Clark
1990, Wursig and Clark 1993, Matthews et al. 2001, Laurinolli et al. 2003) in
previous publications. Although the Gunshot sound is acoustically similar to other
impulsive sounds, such as the sounds produced from flipper slaps or lobtails
(Watkins 1981, Wiirsig and Clark 1993), it appears to have an internal production
mechanism in right whales, with no externally visible movements of flippers, flukes
or jaws during the generation of the sound.
There are no published reports of behavior related to the production of this sound
in North Atlantic right whales. Earlier results (Parks 2003a), and observations
described in this paper, indicate that Gunshot sounds do not attract other right
whales. Here we describe Gunshot sounds recorded from North Atlantic right
whales in two behavioral contexts. These include stereotyped displays by lone adult
males and surface active groups (SAGS) (Kraus and Hatch 2001) with multiple
whales present.

METHODS

Behavioral Data Collection
Data were collected during summer and autumn vessel-based right whale surveys
conducted by the New England Aquarium between 1990 and 2002 and surface active
group behavioral observations carried out by S. Parks from 1999 to 2002 in the Bay
of Fundy, Canada. The New England Aquarium surveys have been conducted since
1980 in the Bay of Fundy and focused on the identification and distribution of
individual right whales in their summer and autumn habitat. Only data from 1990
to 2002 were included in this paper to reflect the survey period with the most
consistent documentation of acoustic recordings and behavioral observations. When
a right whale was sighted, the survey vessel would approach the whale to collect

460

zyxwvutsr
zyxwvu
zyxwv
zyxwvuts
MARINE MAMMAL SCIENCE, VOL. 21, NO. 3 , 2005

photographs, document the position of the whale, and then return to the survey
transect. During these surveys, additional behavioral data were collected
opportunistically, including the documentation of associations between individuals,
and video and audio recordings of any behaviors considered noteworthy.
The primary behavioral interaction observed in North Atlantic right whales in
the Bay of Fundy is the surface active group (SAG). These groups are defined as two
or more animals interacting at the surface, less than one body length apart with
frequent physical contact (Kraus and Hatch 2001). The animal at the center of the
SAG around which most of the activity appears to be focused is referred to as the
focal whale. These focal whales frequently roll upside-down with their bellies
exposed at the surface in the center of the group. The focal whale in groups of three
or more animals in the Bay of Fundy is typically female while other whales in the
groups are generally males (Kraus and Hatch 2001).
Behavioral observations and acoustic recordings of SAGS were made by S. Parks
from 1999 to 2002. This study involved finding right whales in SAGS and staying
with the group to observe it for as long as possible. Once a group was located,
photographs were taken to determine which individuals were involved in the
group. Acoustic recordings were made of the sounds produced in the group. Both
single hydrophone recordings (1999-2001) and 15-element hydrophone array

recordings (2002)were made. Behavioral data collected included changes in group
composition resulting from arrivals and departures of individual whales, the
identity of the focal animal and the distance of the groups from the recording
platform. Recordings of Gunshot sounds from lone whales and SAGS were made
during both of these studies.

Individual ldentifcation
Photographs were taken of head callosities, flukes, and any distinguishing body
scars of whales to use for individual identification as described in Kraus et al. (1986).
Photographs were compared to the North Atlantic right whale catalog (Hamilton
and Martin 1999) by the New England Aquarium right whale research group in
Boston, MA, for identification of individuals. The catalog of individuals includes
information on the sex of individuals in the population based on direct observation of
the genital area, long-term association with a newborn calf, and genetic typing of
tissue samples (Brown et al. 1994). Life histories, including age and mother-offspring
relatedness are known for most individuals born into the population since 1980
(Hamilton and Martin 1999). This database allowed us to determine the sex and age
for the whales observed in this study. Many of the whales in the catalog were first
observed as juveniles or adults, so their exact age is unknown. The estimate of age was
made from the year of first sighting to provide a minimum estimate of the age of the

animals. For example, a whale listed as 12+ yr of age had a 12-yr sighting history and
was first sighted as a juvenile or adult. Many of the whales with minimum estimates
of age may be much older than the minimum ages reported here. Although little is
known about the age of sexual maturity in male right whales, the average age of first
parturition for females is 9.5 yr (Kraus et al. 2001). For this paper, males were
considered to be adult at 9 yr or older.

zyxwvu

Acoustic Data Collection
From 1990 to 1998, single hydrophone acoustic recordings were made with
a Sonotronics hydrophone and a Marantz Model PMD430 recorder with a flat

zyxwvutsrq
zyxwvutsr
zyxwv

PARKS ET AL.: RIGHT WHALE COMMUNICATION

46 1

zyxw

frequency response from 35 Hz to 17 kHz. Video of behaviors from 1992 to 1998
were made using Sony Hi8 Handycam (Model CCD-TR81). From 1999 to 2002,
single hydrophone acoustic recordings were made with an Hi-Tech HTI-94-SSQ
hydrophone (frequency response 50 Hz-20 kHz, k 0 . 5 dB) and a TASCAM DA-P1
DAT recorder (nominal frequency response 20 Hz-20 kHz, k0.5 dB). Additional
acoustic recordings were made with the HTI hydrophone connected directly to
a SONY DCR-TRV V900 MiniDV camera to provide synchronized audio and
video recordings of the observed behaviors. In the single hydrophone recordings of
lone whales, no other whales were visible in the immediate area during the
recording period and the recorded sounds were of high intensity, suggesting that
the observed individual produced all of the sounds.
In 2002 an array recording system was used to determine direction to a recorded
sound consisting of a 3.75-m rigid linear array of 15 Benthos AQ-2TS hydrophones
with custom preamps with 40 dB gain. The array design and pre-amps were based
on an array design developed for use with killer whales (Orcinzs orca) (Miller and
Tyack 1998). The hydrophones were evenly spaced 0.25 m apart. The array was
manually deployed at 5 m depth off the side of a stationary 7-m outboard vessel, the

R/V Callisto in 2002. The array was originally designed for towing behind a vessel
in motion. For this study the array was buoyed with foam spar buoys made of pipe
insulation in three evenly spaced points along the length of the array. This
maintained a horizontal position of the array in the water while the vessel was
stationary during behavioral observations. The signal from each hydrophone was
acquired through a low power (30 mW), 15-channel, computer programmable,
bandpass filter with adjustable corner frequencies and gain designed by Robert B.
MacCurdy, an electrical engineer at the Bioacoustic Research Program at Cornell
University (CUBRP). The digital data acquisition set up consisted of a National
Instruments PCMCIA DAQCard-6062E (12-bit 500 kS/s sampling) in a Dell
Latitude C6lO laptop using the program Chickadee developed at CUBRP. Sounds
for all channels were acquired at 8 kHz sampling. The beamforming frequency
cutoff was 2.5 kHz. Near real time beamforming was carried out in the field to
determine the bearing to sound sources during behavioral observations. The
CUBRP Matlab Time-Delay Beamformer used acoustic data from four hydrophones
in the array and was developed by Kathy Dunsmore and Kurt Fristrup at Cornell
University. The same software was used for later processing of saved files to confirm
that observed bearings to whales agreed with the calculated bearings to the sounds.

Somd Analysis

Sound analysis was done using Cool Edit Pro (Syntrillium, Adobe). The
characteristics of the signal and the rates of production were investigated from all
recordings with a high signal-to-noise ratio. Measurements were taken of the
number of broadband pulses present in the signal, the timing between pulses and
the time between the onset of the sound and the recorded echo off the bottom.
Gunshot sounds recorded from a whale in a known position were analyzed to
estimate the source level. The position of the whale was determined using the
beamformer to determine bearing to the sound source and using a Leica LRG 800
laser range finder to measure range to the whale at that bearing at the surface.
Received level measurements were made using custom wrirten scripts in Matlab 6.5
(Mathworks, Natick, MA, USA). The received level at che Hi-Tech hydrophone
(sensitivity -170 dB re: 1V/1 pPa) was calculated relative to a calibration signal

462

zyxwvu
zyxwv
zyxwvut
zyxwvutsr
zyxwvuts
MARINE MAMMAL SCIENCE, VOL. 21, NO. 3 , 2005

with a known 3 kHz RMS voltage level recorded directly into the DAT recorder.
Peak-to-peak (p-p) sound pressure level (dB p-p re: 1pPa) was given by the p-p
amplitude difference between the signal relative and the p-p value of the calibration
signal. The RMS sound pressure level (dB rms re: 1pPa) was calculated by taking
the root of the mean pressure squared in a time window T. This duration T was
defined by the sample fraction that generates 90% of the total cumulative energy in
a window including the sound pulse (Madsen et al. 2004). Source Level (SL) and
Transmission Loss (TL)were estimated for the R L of recorded calls by SL = RL
TL, where TL = 20 lo&) under the assumption of spherical spreading (Urick
1983). Absorption had negligible effect (